Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    the word "Think" at the beginning of phrases, likely indicating prompts for reflection or consideration

    oai_token-act-pair · gpt-3.5-turbo

    instances where the word "Think" is prompted to encourage reflection or consideration on various topics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    cffffcc
    -0.82
    onduct
    -0.72
    SPONSORED
    -0.71
    catentry
    -0.69
    taboola
    -0.68
    ategory
    -0.66
    artney
    -0.63
    rypt
    -0.63
    ORPG
    -0.63
     camer
    -0.63
    POSITIVE LOGITS
    stock
    0.80
    pad
    0.73
    about
    0.69
    table
    0.69
    Pad
    0.68
     about
    0.67
     Twice
    0.67
     Rice
    0.66
    progress
    0.64
    ventus
    0.64
    Act Density 0.044%

    No Known Activations